Constraints on the Sum of Neutrino Masses from ACT DR6 and DESI DR2 Considering Isocurvature Initial Conditions

This paper presents the first joint constraints on the sum of neutrino masses and neutrino density isocurvature modes using Planck 2018, ACT DR6, SPT-3G, DESI DR2, and DES Year 5 data, finding that while the inclusion of isocurvature perturbations only marginally weakens the upper mass limits, the tightest bounds are highly sensitive to dark energy models and prior assumptions regarding the neutrino mass hierarchy.

The Gist

Imagine the universe as a giant, expanding balloon. For decades, scientists have been trying to figure out how heavy the "invisible passengers" on this balloon are. These passengers are neutrinos—tiny, ghost-like particles that zip through everything without interacting much. Knowing their total weight (the sum of their masses) is a huge puzzle in physics.

This paper is like a team of detectives using the most powerful telescopes and surveys available (ACT, DESI, DES, and Planck) to weigh these neutrinos. But there's a catch: to weigh them, the detectives have to make some assumptions about how the balloon started inflating.

Here is the story of what they found, broken down into simple concepts:

1. The Two Ways the Universe Could Have Started

Usually, scientists assume the universe started with a "perfectly smooth" beginning, like dough rising evenly in a pan. This is called Adiabatic initial conditions. In this scenario, everything (light, matter, neutrinos) started moving together in perfect sync.

However, the authors asked: What if the universe started a bit "messy"? What if the neutrinos started with their own separate rhythm, out of sync with the rest of the universe? This is called Isocurvature (or NDI). It's like if the dough had little pockets of yeast that started rising at different speeds.

2. The Big Question

The main question of the paper is: If we allow for this "messy" start (Isocurvature), does our estimate of the neutrinos' weight change drastically?

If the answer is "Yes," then our current weight limits are fragile and depend too much on our assumptions. If the answer is "No," then our weight limits are solid and reliable, no matter how the universe started.

3. The Investigation

The team combined data from:

• The Cosmic Microwave Background (CMB): The "baby picture" of the universe (from Planck, ACT, and SPT-3G).
• Galaxy Surveys (DESI and DES): Maps of how galaxies are spread out today.

They ran two simulations:

1. Scenario A: The universe started perfectly smooth (Standard).
2. Scenario B: The universe started with a "messy" neutrino rhythm (Isocurvature allowed).

4. The Results: The Weight Limit Holds Firm

Here is what they discovered, using a simple analogy:

Imagine you are trying to guess the weight of a hidden object in a box.

• In the "Smooth" scenario (Standard): They calculated the object weighs less than 0.052 eV (a tiny, tiny amount).
• In the "Messy" scenario (Isocurvature): They recalculated allowing for the messy start. The limit shifted slightly to 0.057 eV.

The Verdict: The limit barely moved! The "messy" start didn't throw off the scale. The data shows no evidence that the universe started with this messy neutrino rhythm. The "messy" component is consistent with zero.

Why does this matter? It means the current upper limit on neutrino mass is robust. Even if the universe started slightly differently than we thought, our conclusion that "neutrinos are very light" remains true.

5. The Twist: The "Dark Energy" Factor

While the neutrino weight limit was stable against the "messy start," it was very sensitive to a different assumption: Dark Energy.

Dark Energy is the mysterious force pushing the universe apart.

• If Dark Energy is a constant (like a steady wind), the neutrino limit is tight (< 0.052 eV).
• If Dark Energy changes over time (like a wind that speeds up or slows down), the limit loosens significantly to < 0.111 eV.

The Analogy: Think of the neutrino weight limit as a rubber band.

• Changing the "start of the universe" (Isocurvature) barely stretches the rubber band.
• Changing the "nature of Dark Energy" stretches the rubber band by 50%.

This tells us that to get a perfect, final answer on neutrino mass, we need to understand Dark Energy better than we do right now.

6. The "Floor" Problem

There is one final interesting detail. The paper notes that their tightest limit (0.052 eV) is actually lower than the minimum weight neutrinos must have based on what we know from particle physics (the "Normal Hierarchy" requires at least 0.05878 eV).

This is like a scale saying a person weighs 100 lbs, but we know for a fact they can't weigh less than 120 lbs. The paper explains this isn't a physical reality; it's a statistical artifact caused by the math allowing the weight to go down to zero. When they fix the math to respect the known minimum weight, the limit becomes 0.092 eV.

Summary

• Did the "messy start" break the results? No. The neutrino mass limit is very stable even if the universe started with a different rhythm.
• Did we find a "messy start"? No. The data suggests the universe started smoothly.
• What is the biggest uncertainty? Our understanding of Dark Energy. If Dark Energy changes over time, the neutrino mass limit becomes much looser.
• Conclusion: Current data gives us a very strong, reliable upper limit on how heavy neutrinos can be, provided we accept that Dark Energy might be a bit more complicated than a simple constant.

Technical Summary

Technical Summary: Constraints on the Sum of Neutrino Masses from ACT DR6 and DESI DR2 Considering Isocurvature Initial Conditions

Problem Statement
The absolute mass scale and mass hierarchy of neutrinos remain fundamental open questions in modern physics. While laboratory experiments (e.g., KATRIN, neutrinoless double-beta decay) provide upper limits, cosmological observations offer significantly tighter constraints on the sum of neutrino masses ($\sum m_\nu$). However, these cosmological bounds are inherently model-dependent, relying heavily on the assumption of purely adiabatic primordial initial conditions. Standard $\Lambda$CDM analyses assume all species share the same curvature perturbation. This paper addresses the robustness of current $\sum m_\nu$ constraints when relaxing this assumption to allow for a neutrino density isocurvature (NDI) component, where the photon-to-neutrino density ratio varies spatially. The authors investigate whether the inclusion of NDI modes significantly weakens the upper limits on neutrino mass derived from the latest high-precision datasets.

Methodology
The authors perform a joint Bayesian analysis using the most recent cosmological datasets:

• CMB Data: A combination of Planck 2018, Atacama Cosmology Telescope (ACT) Data Release 6 (DR6), and South Pole Telescope third-generation (SPT-3G) data, referred to as the "CMB-SPA" likelihood. This combination leverages ACT's high-resolution small-scale temperature/polarization maps and SPT-3G's deep polarization sensitivity to break degeneracies between $\sum m_\nu$ and parameters like $\Omega_m$ and $H_0$.
• Large-Scale Structure (LSS): Baryon Acoustic Oscillation (BAO) measurements from DESI Data Release 2 (DR2), covering over 14 million tracers, and Type Ia supernova (SN) data from the Dark Energy Survey (DES) Year 5.

Theoretical modeling involves:

1. Initial Conditions: The study compares two scenarios: (i) a standard purely adiabatic model, and (ii) a mixed model allowing for uncorrelated NDI modes. The primordial power spectra are parameterized using amplitudes at two pivot scales ($k_1 = 0.002 \, \text{Mpc}^{-1}$ and $k_2 = 0.1 \, \text{Mpc}^{-1}$) to capture signatures on both large scales (CMB acoustic peaks) and small scales (damping tail/matter power spectrum).
2. Dark Energy Models: Analyses are conducted within the standard $\Lambda$CDM framework and the Chevalier–Polarski–Linder (CPL) dynamical dark energy model ($w(a) = w_0 + w_a(1-a)$) to assess sensitivity to the dark energy equation of state.
3. Neutrino Hierarchy: Constraints are derived under three hierarchy assumptions: Normal Hierarchy (NH), Inverted Hierarchy (IH), and Degenerate Hierarchy (DH), as well as under different prior lower bounds for $\sum m_\nu$ (ranging from 0 to the minimum mass required by the hierarchy).
4. Inference: Parameter inference is performed using MontePython interfaced with the CLASS Boltzmann solver, employing Markov Chain Monte Carlo (MCMC) sampling.

Key Contributions

• First Joint Constraint: This work provides the first joint constraint on $\sum m_\nu$ and NDI amplitudes using the full CMB-SPA + DESI DR2 + DES SNe dataset.
• Robustness Assessment: It systematically quantifies how the inclusion of NDI modes affects the inferred neutrino mass limits, moving beyond the standard adiabatic assumption.
• Prior Dependency Analysis: The paper explicitly highlights the sensitivity of the derived upper limits to the assumed prior lower bound on $\sum m_\nu$, particularly in the context of the neutrino mass hierarchy.

Results

• $\Lambda$CDM Model:
  • Under purely adiabatic initial conditions with a prior lower bound of 0 eV, the 95% upper limit is $\sum m_\nu < 0.052$ eV.
  • When allowing for NDI modes, the limit weakens only marginally to $\sum m_\nu < 0.057$ eV.
  • The NDI amplitude is consistent with zero at 95% confidence, indicating no evidence for a non-adiabatic component.
  • The authors note that the 0.052 eV limit lies below the minimum mass required by the Normal Hierarchy ($0.05878$ eV). When imposing a physically motivated lower bound ($\sum m_\nu \geq 0.05878$ eV), the 95% upper limit rises to $< 0.092$ eV (adiabatic) and $< 0.094$ eV (with NDI).
• CPL Dynamical Dark Energy Model:
  • The upper limit loosens significantly to $\sum m_\nu < 0.111$ eV (adiabatic) and $< 0.115$ eV (with NDI), demonstrating high sensitivity to the dark energy equation of state.
  • Despite the relaxation in absolute values, the inclusion of NDI modes still results in only a marginal shift ($\sim 0.004$ eV), and the NDI amplitude remains consistent with zero.
• Degeneracy Breaking: The distinct scale-dependent signatures of massive neutrinos (suppressing small-scale power) and NDI modes (affecting large-scale CMB peaks and intermediate-scale matter power) allow current high-precision data to break the degeneracy between $\sum m_\nu$ and the NDI amplitude.

Significance and Claims
The paper claims that current cosmological data show no evidence for neutrino isocurvature modes and that the inferred upper limits on the sum of neutrino masses are robust against the inclusion of NDI extensions. The authors emphasize that the stability of the bounds is not merely a lack of sensitivity but a positive indication that the data favor purely adiabatic initial conditions.

However, the authors modestly conclude that a definitive, model-independent bound on $\sum m_\nu$ requires addressing two major uncertainties:

1. Prior Dependencies: The absolute value of the upper limit is highly sensitive to the assumed lower bound, particularly regarding the neutrino mass hierarchy. They recommend reporting limits with both a zero lower bound (for comparison) and a hierarchy-informed lower bound (for physical plausibility).
2. Dark Energy Uncertainties: The inferred mass scale is strongly dependent on the assumed dark energy equation of state, with the CPL model yielding limits roughly 50% weaker than $\Lambda$CDM.

The work serves as a crucial benchmark for future surveys (e.g., CMB-S4, Euclid, LiteBIRD), suggesting that while current data constrain NDI modes effectively, future experiments will be required to definitively detect or rule out isocurvature components and achieve a model-independent determination of the absolute neutrino mass scale.